Dear Haoran,
2011/3/10 飞鸟 <[log in to unmask]>:
> Dear SPM's users,
> When we construct a specific DCM model, we should specify different kinds
> of connections to those sources, but:
> 1. I wonder that whether the lateral connections only exist between the
> sources which are in the same level ? And, what does ' the same level ' mean
Typically, lateral connections are for connecting bi-lateral symmetric
sources i.e. for modelling inter-hemispheric connections. There are
some other cases when they might be used, when it is known that two
areas are at the same level of cortical hierarchy. Usually such
information is available for the well-studied visual system.
> ? Does it mean the two regions own the same function, such as the left
> visual region and the right visual region ?
That's one example.
> 2. Does forward connections only exist between the sources which are in
> the different level ? Such as low-level and the high-level ?
> 3. Do backward connections are always generate from high-level regions and
> terminate in those low-level regions ?
Yes, that's the idea.
> All in a word, how could we specify the different connections among those
> sources ? Any help will be appreciated.
>
A principled way of doing this is model comparison. Of course if you
want to try the 3 possible options for each connection, you'll
probably have very large number of possibilities. But typically you
would have information about at least some of the connections from the
literature. So if you are not sure about a particular connection, you
can try all the relevant options and do model comparison. Note that if
there is bi-lateral symmetry in your model it wouldn't make sense to
assign different types to symmetric connections in the two
hemispheres. This further constraints your options.
Best,
Vladimir
> --
> Haoran LI (MS)
> Brain Imaging Lab,
> Research Center for Learning Science,
> Southeast University
> 2 Si Pai Lou , Nanjing, 210096, P.R.China
>
>
|